Personalized Federated Learning for ECG Classification Based on Feature Alignment

Personalized Federated Learning for ECG Classification Based on Feature Alignment
复制标题

基于特征对齐的心电分类个性化联邦学习

DOI:
10.1155/2021/6217601
复制
发表时间:
2021-11
期刊:
Security and Communication Networks (SCN)
影响因子:
--
通讯作者:
Jiahui Jin
Jiahui Jin
中科院分区:
其他
文献类型:
--
作者:
Renjie Tang;Junzhou Luo;Junbo Qian;Jiahui Jin

文献摘要

参考文献

相似文献

心电图数据分类是医学信息处理中的一个研究热点。然而,数据不足,隐私保护和本地部署仍然是具有挑战性的困难。针对这些问题,本文提出了一种新的个性化联邦学习心电图分类方法。首先,在多个本地数据客户端上使用联邦学习框架训练全局模型。然后,我们使用全局模型和私有数据来训练局部模型。为了减少全局数据和局部数据之间的特征不一致性,更好地拟合局部数据,设计了一种新的“特征对齐”模块来保证一致性,该模块分为全局对齐和局部对齐两部分。对于全局对齐,使用批数据的图度量来约束全局模型和局部模型生成的特征之间的不相似性。对于局部对齐,采用三元组丢失来增加对局部私有数据的区分能力。对我们收集的数据集进行了全面的实验评估。结果表明,该方法能更好地适应局部数据,并具有上级的泛化能力。
Electrocardiogram (ECG) data classification is a hot research area for its application in medical information processing. However, insufficient data, privacy preserve, and local deployment are still challenging difficulties. To address these problems, a novel personalized federated learning method for ECG classification is proposed in this paper. First, a global model is trained with federated learning framework on multiple local data clients. Then, we use the global model and private data to train the local model. To reduce the feature inconsistency between global and private local data and for better fitting the private local data, a novel ”feature alignment” module is devised to guarantee the uniformity, which contains two parts, global alignment and local alignment, respectively. For global alignment, the graph metric of batch data is used to constrain the dissimilarity between features generated by the global model and local model. For local alignment, triplet loss is adopted to increase discriminative ability for local private data. Comprehensive experiments on our collected dataset are evaluated. The results show that the proposed method can be better adapted to local data and exhibit superior ability of generalization.
DOI: --
发表时间: 2020-02
期刊: ArXiv
影响因子: --
作者:
Y. Mansour;M. Mohri;Jae Ro;A. Suresh
通讯作者: Y. Mansour;M. Mohri;Jae Ro;A. Suresh
DOI: 10.1016/j.physa.2003.12.021
发表时间: 2004-05
影响因子: 3.3
作者:
R. Muirhead;R. Puff
通讯作者: R. Muirhead;R. Puff
DOI: 10.1016/j.dsp.2006.11.009
发表时间: 2007-05
期刊: Digit. Signal Process.
影响因子: --
作者:
Elif Derya Übeyli
通讯作者: Elif Derya Übeyli
DOI: 10.1609/aaai.v35i9.16960
发表时间: 2020-07
期刊: --
影响因子: --
作者:
Yutao Huang;Lingyang Chu;Zirui Zhou;Lanjun Wang;Jiangchuan Liu;J. Pei;Yong Zhang
通讯作者: Yutao Huang;Lingyang Chu;Zirui Zhou;Lanjun Wang;Jiangchuan Liu;J. Pei;Yong Zhang
DOI: --
发表时间: 2016-10
期刊: ArXiv
影响因子: --
作者:
Jakub Konecný;H. B. McMahan;Daniel Ramage;Peter Richtárik
通讯作者: Jakub Konecný;H. B. McMahan;Daniel Ramage;Peter Richtárik